Part II Laboratory Studies L829/frame/ch07 Page 157 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC 159 7 LAMMA and Raman Study of Oxidation States of Chromium in Aerosols: Application to Industrial Hygiene A. Hachimi, E. Poitevin, G. Krier, and J.F. Muller CONTENTS Introduction 159 Study of Chromium in Polyphasic Dust 160 Materials and Methods 160 Sampling 160 Analytical Techniques 161 SEM Analysis 161 Raman Microprobe 161 X-ray Photoelectronic Spectroscopy (XPS) 161 Laser Microprobe Mass Analysis (LAMMA) 161 Analysis of Dusts from Welding Fumes 162 Qualitative Analysis 162 Calculation of Chromium Amount with Different Stoichiometry 164 Fibrous Aerosols 164 Microspherical Aerosols 164 Complementary Analysis 165 Determination of Chromium Valency in Aerosols Less than 10 µ m in Size Emitted as Dust from the Steel Industry 166 Elemental Analysis of Dust Collected from the Site 167 Study of the Valency of Chromium in Dusts Collected by a Portable Impactor and by a Normal Andersen Impactor at the Same Site 168 Chromium Analysis of Large Dust Aerosols Collected with the Portable Impactor 168 Conclusion 174 References 174 INTRODUCTION Since its discovery in 1797, chromium has been increasingly used in industry. Its effects on human health have been gradually understood. The two most important sights of professional pathology, due to chromium and to its derivatives, are represented by “allergogene” action and the cancer- producing power of some products. Indeed, hexavalent chromium derivatives cause dermatosis of contact, bronchitic asthma, perforation of the nasal septum, and bronchiopulmonary cancers. 1,2 L829/frame/ch07 Page 159 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC 160 Aerosol Chemical Processes in the Environment Experimental studies have shown that chromates and bichromates are able to induce in vitro and in vivo cancers in animals. 3-5 Epidemiologic survey have shown that the lungs represent a target organ of hexavalent derivatives of chromium. However, only a few experimental studies exist that have allowed for the thorough study of toxicity mechanisms. In welding and steel working, high concentrations of fumes and gas are emitted, which contain chromium in either the hexavalent or trivalent form; thus, health problems in relation to the presence of chromium in dust fumes can occur. Therefore, methodology application on valency determination of chromium in environmental dusts was desirable. In the literature, one finds two kinds of techniques for the determination of chromium valency in aerosols of industrial origin: • Chemical spectrophotometric analysis techniques, also called wet chemical techniques • Physical and chemical techniques using direct measurement and spectroscopic techniques Wet chemical techniques, such as colorimetry, ion exchange resins, luminescence, and atomic absorption, 6-8 have the ability to determine both chromium identification and correct proportioning in its numerous oxides forms. A main drawback concerns aerosol sampling because the sample can be physically altered (e.g., by extraction, dissolution, electrochemical reactions during lixivi- ation operation, etc.) with the possibility of induced chemical changes, matrix effects, and differ- ential solubility artifacts. 9 Spectroscopic techniques such as Raman, X-ray diffraction, FTIR, 10,11 XPS, 12-14 and XRF 15,16 allow in situ characterization of solid aerosols, but there are many instrumental limitations related to detection, sensitivity, sampling, and data interpretation. Among the spectroscopic techniques recently described in the literature to identify in situ element oxidation rates, mass spectrometry presents new insight on this problem, especially laser microprobe LAMMA. One can apply the valency determination method for chromium to a complex matrix: on the one hand, arc welding fumes on stainless steel (MMA/SS) and, on the other hand, dust from the steel industry. Thus, one can investigate the determination of major oxidation rates of chromium derivatives contained in the dust that is directly inhaled by workers. STUDY OF CHROMIUM VALENCY IN POLYPHASIC DUST The method proposed here involves the determination of chromium valency, in aerosols less than and greater than 10 µ m in diameter, emitted by welding and in aerosols, 0.4 to 10 µ m in diameter from various steel works. This method has been improved upon by precise sampling with granu- lometric discrimination of aerosols (using cyclone and Andersen impactors) and targeting of fume emission sites at steel-making locations. Microprobe Raman techniques, and ESCA and SEM, have been used as complementary techniques of confirmation. MATERIALS AND METHODS S AMPLING Welding dust was collected after a welding operation on stainless sheet steel (MMA/SS) of type 304L 18-10 (lower carbon) with rods AROSTA — basic rutile electrode 304. Dusts pass in a cyclone impactor by suction, drawing up a flow of 1.71 l min –1 for 46 min. Dusts are finally collected onto a nitrocellulose filter of 0.2 µ m porosity and 32-mm diameter. Sampling selection of dust was fixed to 10 µ m coating element of the rods and trimming dusts (metallic projectiles) also have been collected in the experimentation workshop for LAMMA analysis. L829/frame/ch07 Page 160 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC LAMMA and Raman Study of Oxidation States of Chromium in Aerosols 161 Dust sampling was performed at a site rich in fumes containing large amounts of chromium using two kinds of impactors: • A normal Andersen cascade impactor with an air flow rate of 28 l min –1 for discrimination of aerosols in air flux by inertial impaction relative to their mean aerodynamic diameter. Size fractionation in nine stages (>9 µ m; 9 to 5.8 µ m; 5.8 to 4.7 µ m; 4.7 to 3.3 µ m; 3.3 to 2.1 µ m; 2.1 to 1.1 µ m; 1.1 to 0.7 µ m; 0.7 to 0.4 µ m; and <0.4 µ m) gave information about the distribution of chromium compounds in relation to aerosol size. • Dusts were collected by impaction on aluminium filters during a period of 12 to 24 h; a portable cascade impactor consisting of five stages, with aerosols size ranging from 0.4 to 9.8 µ m and air flow rate of 1.7 l min –1 carried by workers during an 8-h shift. Dusts were collected on nitrocellulose or polyethylene terephthalate filters. A NALYTICAL T ECHNIQUES SEM Analysis Element analysis was achieved using a JEOL 840 electron microscope coupled to a dispersive energy spectrometer. An electron beam current of 200 to 300 pA and accelerating voltage of 15 kV were used. Raman Microprobe MicroRaman analysis of aerosols collected on the portable cascade impactor was performed on a standard DILOR XY apparatus equipped with an argon ion laser and multichannel detector (1024 diodes). Excitation wavelength was 514.5 nm using a power of 75 mW to avoid fluorescence emission that could mask Raman diffusion. Spectral resolution was 4 cm –1 , and integration time varied from 1 to 15 s. The spectrometer was coupled to an optical microscope (Olympus), permitting a spatial reso- lution of 2 µ m. Spectra were obtained in reflective mode and presented in arbitrary units vs. wavenumbers (cm –1 ). X-ray Photoelectronic Spectroscopy (XPS) The used apparatus from Leybold Heraeus TM had a resolution of 1 eV. The radiation source comes from the K α radiation of aluminum. The deposited dusts are directly mounted on the sampling support. An interval of 3 eV, on average, has been recorded for the characteristic lines of C(1 s ), K(2 p ), and K(2 s ), O(1 s ), and F(1 s ). This interval is due to the charge effects from fibrous dusts and the nitrocellulose filter. Laser Microprobe Mass Analysis (LAMMA) LAMMA was developed for localization and determination of elements in various samples — either conducting or isolating ones. An interesting feature of this technique is its ability to charac- terize the molecular composition of inorganic substances. Moreover, it allows for elementary analysis without the traditional separation step. Element detection limits are 10 –15 to 10 –19 g. This sensitivity allows for LAMMA analysis of aerosols in biological and environmental studies. 17,18 The study of ionized clusters in both positive and negative modes is correlated with the sample chemical composition (e.g., SO – , SO 2 – , SO 3 – , SO 4 – , NaSO 3 – , NaSO 4 – ). Major ions obtained by laser ionization are representative of sulfate and sodium thiosulfates. 19 The information is useful to complete data obtained by other techniques. Additionally, a link between morphological properties of particles can be established by sample observation under visible light. L829/frame/ch07 Page 161 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC 162 Aerosol Chemical Processes in the Environment Technological progress (laser, configuration of ionization chamber) has permitted increased use of this method, and has allowed for the analysis of organic and inorganic matter ((nitro)PAH desorption) and matrix identification. 20,21 The laser microprobe already has a privileged place and ongoing instrumentation progress will make certain its success. Impacted dusts were extracted from filters by superficial scraping and set by simple pressure on a microscopic grid coated with a formvar film. Comparisons could be made with constant instrumental parameters, and the LAMMA apparatus had the following features: • Wavelength: 266 nm • Pulse width: 12 ns • Laser focus: 2–3 µ m • Energy on sample: 1–3.5 µ J • TOF voltage: ±3000 V • Extraction lens voltage: ±1000 V • Reflection voltage: ±790 V • Cathode voltage: ±6000 V All spectra were recorded on Nicolet 4094C, a recorder connected to an Apple Macintosh II CX computer with a 40-Mbyte system. Spectra were calibrated and linearized following mass spectrometry conventions. ANALYSIS OF DUSTS FROM WELDING FUMES Welding fumes can contain large amounts of chromium compounds with varying concentrations that depend on the welding process and the rod composition. A systematic study was achieved on these different elements for the following: • Coating of the rod (AROSTA, rutile basis electrode type 304 L) • Trimming dusts (metallic projectiles) • Welding aerosols less than 10 µ m in diameter that have metallic microspherical form and are collected in the impactor after welding operation • Welding aerosols more than 10 µ m in diameter that are microfibers collected onto nitrocellulose filters Q UALITATIVE A NALYSIS We have investigated fingerprint spectral analysis of the different elements cited above. LAMMA spectra are presented in Figures 7.1 to 7.5. General observations of these spectra lead to to the following remarks: • The constitutive elements of the coatings of the rod, such as Na, K, Ca, F, Cl, Mn, P, and Si (in aluminosilicate form), are present in all the aerosols. • Barium, which is present in its oxide and fluoride forms and comes from coatings of the rod, is simply present in the microspherical aerosols (>10 µ m). • Trimming compounds, likely microspheres, contain metals (Cr, Al, Ti, and Ni) that are representative of the elemental composition of the rod and stainless steel sheet-metal. • Fibrous dusts (<10 µ m) with a basic nature (fluorides and chloride compounds) likely contain an alumino- and ferrosilicate matrix in which metallic elements are scattered. • Microspheres contain a lot of sulfate compounds, like trimming compounds, character- ized by SO x – clusters in negative ionization mode, whereas microfibrous aerosols do not. L829/frame/ch07 Page 162 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC LAMMA and Raman Study of Oxidation States of Chromium in Aerosols 163 • The two different spectra of fibrous aerosols in positive mode show the heterogeneity of the matrix, one can obtain either spectra with metallic nature of spectra with alumino- and ferrosilicates matrix. • Chromium and nickel are systematically present in positive spectra (Cr + and Ni + ) in both kinds of aerosols. The entire LAMMA analysis is presented in Table 7.1 where elemental ions and combinations present in plasma are displayed. TABLE 7.1 Elemental Analysis of Welding Dust by LAMMA Element Rod Coating Trimming Dust (>10 µ m) Dust (<10 µ m) Si +++ +++ ++ +++ +++ Mn + + + + ++ Cr ++ – ++ +++ ++ Ni ++ – ++ ++ Al – ++ ++ ++ ++ O – ++ ++ ++ +++ F – ++ +++ +++ +++ Na – +++ ++ ++ +++ Cl – ++ ++ ++ ++ K – +++ +++ +++ +++ Ca – +++ + +++ + Fe – ++ +++ ++ + S – ++ ++ – Ti –– ++ +++ + Mg –– ++ + Ba – ++ – +++ – Type of Cluster Ions Present on Positive and Negative Modes Al x O y +++ ++ +++ + Si x O y +++ ++ +++ +++ PO x ++ +++ + +++ Fe x O y +++ ++ + AlSi x O y +++ ++ + ++ CaF + + +++ + SO x +++ ++ – FeSi x O y ++ –– CrO x – ++ ++ ++ BaOH ++ – ++ – TiO – + +++ + BaF –– + – K x F y + –– +++ Na x Cl y –– – ++ CaCl –– – + K x Cl y –– – + Al(Na x Cl x ) –– – ++ Na(K x F x ) –– – ++ K(Na x Cl x ) –– – ++ Na 2 OH –– – + K 2 OH –– – + Note : Absent: –; Present: +; Majority: ++; Abundant: +++ L829/frame/ch07 Page 163 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC 164 Aerosol Chemical Processes in the Environment C ALCULATION OF C HROMIUM A MOUNT WITH D IFFERENT S TOECHIOMETRY Oxidation state studies were performed on 85 and 24 individual spectra of fibrous and microspher- ical aerosols, respectively. Investigation of 455 and 325 accumulated spectra allow for the deter- mination of the major chromium oxidation state in the fibrous and microspherical dusts, respectively. All results are summarized in Table 7.2. The table displays the CrO 2 – /CrO 3 – ratios of different individual and mean spectra. Fibrous Aerosols: (<10 µµ µµ m) Spectral analysis of fibrous aerosols has shown that chromium was only present in its elementary form Cr + in positive ions, while CrO 2 – , CrO 3 – clusters (and sometimes CrO 4 – ) were present in negative mode. Energy variation and length variation do not interfere on spectra and intensity ratios of negative clusters. Results are presented in Table 7.3 with analysis details of individual spectra. The methodology was applied on fibrous aerosols with no ambiguity and it confers to chromium an oxidation degree of VI (93% of individual spectra have a chromium VI fingerprint). The salt character of fibers and a value of CrO 2 – , CrO 3 – that corresponds to an anhydrous chromium salt could show chromium in chromate form. In fact, this is in agreement with other results 22,23 on similar type dusts, where chromium has been identified in sodium or potassium form. However, LAMMA does not detect the type of cluster K x CrO y + , CrxO y – (x = 2; y = 4, 5, 6); this is due to either the lower sensitivity or other major recombinations in plasma. Therefore, the assumption of chromium VI presence in the chromate form must be verified by other analytical techniques. Microspherical Aerosols (>10 µµ µµ m) After analysis, the 24 individual spectra of studied microspheres (20 µm in diameter) gave the following results: TABLE 7.2 Ratio Calculation of CrO 2 – /CrO 3 – of Welding Dust Type of Dust CrO 2 – /CrO 3 – N Spectra Mode Wavelength (nm) Fibrous 0.218 ± 0,12 85 Individual 225.7 Dust <10 µm 0.155 70 Accumulation 225.7 0.228 200 Accumulation 286.5 0.181 100 Accumulation 286.5 Total: 455 Weighted mean 0.205 Standard deviation 0.03 Variance 0.0008 Microsphericals 0.754 ± 0.194 24 Individual 225.7 Dust >10 µm 0.673 100 Accumulation 225.7 0.543 100 Accumulation 225.7 0.56 25 Accumulation 225.7 0.551 25 Accumulation 225.7 0.629 50 Accumulation 225.7 Weighted mean 0.614 Total: 324 Standard deviation 0.205 Variance 0.042 L829/frame/ch07 Page 164 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC LAMMA and Raman Study of Oxidation States of Chromium in Aerosols 165 • SO x – ions clusters and CrO – chromium are systematically present (100% of spectra), in addition to CrO 2 – and CrO 3 – ions clusters. • Cluster intensity of sulfur does not interfere with the CrO 2 – /CrO 3 – ratio, which is about 0.8 for the aerosol (arithmetic average of 0.75). A CrO 2 – /CrO 3 – ratio of 0.75 might correspond either to hydrated or anhydrous chromium salt. The lack of CrO 4 – ion and simultaneous presence of CrO – and SO x – for 100% of spectra would reveal a chromium presence as anhydrous chromium sulfate, thus in the III form. One cannot conclude definitively about the real state of chromium in microspheres because the Cr 2 O y – series has not been detected. Nevertheless, two other observations from comple- mentary works would show that chromium is in its III form and in the salt and oxide mixture form: • The CrO 2 – /CrO 3 – ratio varies according to studied microsphere (from 10 to 40 µm in diameter, with value between 0.5 and 4, which corresponds to chromium III and is constant for a particular microsphere). • All microspheres have the same composition and spectral fingerprint (systematic presence of CrO – , SO x – , and lack of CrO 4 – ). These two results support the hypothesis of a mixture containing chromium III salts (sulfate compounds, for instance) and chromium III oxide in variable amounts depending on microspheres. Variation of the CrO 2 – /CrO 3 – ratio from 0.5 to 4 could mean that microspherical aerosols contain an increasing amount of chromium III oxide, depending on microspheres. COMPLEMENTARY ANALYSIS Raman analysis was only effective on fibrous aerosols with a minimal energy to avoid absorption or destruction of dusts. Microspheres study is revealed to be unfeasable because of the large absorption of the aerosols, even with minimal energy application. Results of Raman analysis are shown in Figure 7.1. Two specific bands of the CrO 4 2– ion are obtained at 849 and 901 cm –1 and given hexavalent chromium in salt form in fibrous dust. TABLE 7.3 % Chromium VI Calculated in Welding Dust Dust <10 µm CrO 2 /CrO 3 < 0.3 With CrO 4 Without SO x % of Chromium Spectrum calc. Spectrum number 79 12 85 79 Cr VI CrO 2 /CrO 3 > 0.3 With CrO 4 6 Cr III Spectrum number 6 3 93% Cr VI 7% Cr III Accumulation R = CrO 2 /CrO 3 Valency of majority of chromium 455 spectra 0.205 Value of R confer on chromium VI valency More than 90% chromium Dust >10 µm R > 0.3 With CrO 4 With SO x 24 Cr III Spectrum calc. Spectrum number 24 24 24 100% Cr III Accumulation R Valency of majority of chromium 324 spectra 0.614 Value of R confer on chromium III valency Mixture of chromium III and VI L829/frame/ch07 Page 165 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC 166 Aerosol Chemical Processes in the Environment XPS analysis has likely been performed only on fibrous aerosols because of the deficient amount of microspheres (>10 µm) for effective detection. Major elements already detected by LAMMA analysis are found, and Cr(2p) and Cr(3p) bands for chromium and Ni(2p) and Ni(3p) bands for nickel are observed. The presence of the Cr(2p) band at 579.4 eV after correction for the binding energy indicates that chromium is present in its VI oxidation form, whereas no band is detected at 675.4 eV, which is specific for chromium III. So, the band only at 579.4 eV reveals the presence of chromium exclusively in its hexavalent form. Nickel is present in an oxidized form (a 4-eV shift from the metal binding energy value) with two bands at 858.1 and 876.8 eV for Ni(2p) and Ni(3p), respectively. The LAMMA microprobe turns out to be a sufficiently sensitive and rapid technique for the determination of the oxidation state of chromium included in complex matrices like aerosols emitted from industrial environments. However, it appears from this work that dust sampling (i.e., the conditions of granulometric and morphological selection) and high-risk site targeting are important factors to consider in improving the methodology. That is the reason why we have established a sampling strategy that is more refined for the case of the steel industry, which is an important source of dusts containing non-oxidative products. DETERMINATION OF CHROMIUM VALENCY IN AEROSOLS LESS THAN 10 µµ µµ M IN SIZE EMITTED AS DUST FROM THE STEEL INDUSTRY This study deals with the determination of chromium valency in aerosols less than 10 µm in diameter emitted as dust from the steel industry, and with the development of a strategy for controlling the hazards these aerosols represent. Within the context of industrial hygiene, there is great interest in the way the distribution of trivalent and hexavalent states of chromium differ among the various dust-emitting sites. Dust sampling by cascade impactors at a given site does not necessarily reflect the conditions of inhalation of the aerosols by a person working at that site. Sampling can be improved by fixing a portable cascade impactor to a person in the course of his/her normal activities in an area where chromium- rich dusts are found. FIGURE 7.1 Raman spectrum of dust particles smaller than 10 µm. L829/frame/ch07 Page 166 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC LAMMA and Raman Study of Oxidation States of Chromium in Aerosols 167 ELEMENTAL ANALYSIS OF DUST COLLECTED FROM THE SITE The LAMMA analysis of aerosols in size-fractionated dust samples shows that small aerosols mainly consist of potassium and sodium, with relatively little calcium (Figure 7.2), whereas the large dust aerosols are mainly comprised of calcium (Figure 7.3). FIGURE 7.2 LAMMA spectrum of dust particles smaller than 3.3 µm. FIGURE 7.3 LAMMA spectrum of dust particles smaller than 0.4 µm. L829/frame/ch07 Page 167 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC [...]... by the presence of bands in the range 800 to 70 0 cm–1, similar to those of the ion Cr2O72– (Fe-, Ni-, Zn-, Cu-, Co-, Al-, and Cd-Cro4) Raman results of standard chromium compounds and dust aerosols (8 to 10 µm) collected on the portable impactor are summarized in Table 7. 4 and in Figures 7. 9 and 7. 10 The presence of calcium chromate implied by the SEM and LAMMA analyses is confirmed by this study: the. .. see Table 7. 1) © 2000 by CRC Press LLC L829/frame/ch 07 Page 174 Monday, January 31, 2000 3:02 PM 174 Aerosol Chemical Processes in the Environment CONCLUSION Systematic LAMMA analyses of oxidation states of chromium on aerosols less than 10 µm in diameter emitted in fumes from steelworks have proved that the valency of this metal varies according to the size of the aerosols The smallest aerosols ( . impactor. L829/frame/ch 07 Page 169 Monday, January 31, 2000 3:02 PM © 2000 by CRC Press LLC 170 Aerosol Chemical Processes in the Environment intensities less than 0.1 (Figure 7. 8). The investigations by LAMMA in. development of a strategy for controlling the hazards these aerosols represent. Within the context of industrial hygiene, there is great interest in the way the distribution of trivalent and hexavalent. Dust Aerosols Collected with the Portable Impactor 168 Conclusion 174 References 174 INTRODUCTION Since its discovery in 179 7, chromium has been increasingly used in industry. Its effects on